Mod 25: Assessment of other risks Flashcards
Outline:
1. why it is problematic to use quantitative techniques to analyse liquidity risk
2. the main two analysis techniques adopted in practice ©
Problems in using quantitative techniques to analyse liquidity risk Problems include:
1. limited historic data on liquidity crises
2. the degree and nature of every organisation’s exposure to liquidity risk is different (so industry data or other analogues may not be useful).
Main techniques adopted:
1. scenario analysis: −
liquidity risk is assessed by examining scenarios where the cash outflows exceed the available cash at any point in time
2. stress testing: −
examining the effect on liquidity of an extreme event (a stress scenario) or significant change in a key assumption
©
Outline why scenario analysis may be particularly difficult for retail banks
Why scenario analysis may be particularly difficult for retail banks
Predicting cash outflows is particularly problematic for retail banks. Much of a bank’s liabilities will be in the form of deposits from customers who may withdraw their money with little or no notice (possibly subject to a small penalty on the interest due rather than the capital value).
Cash inflows include:
1. revenues / income generated by assets
2. proceeds from the sale of assets
3. drawings upon sources of liquidity (eg issue of new debt or equity).
It is generally difficult to be confident about the cash proceeds that could be generated from the sale of assets (eg due to the possibility that the sale is forced, or made during a time of depressed asset prices). Retail banks have a particular difficulty, as much of a bank’s asset base is in the form of illiquid long-term mortgages that are not readily convertible into cash.
©
State a particularly important issue in modelling liquidity risk (other than from cash inflows and outflows)
Modelling liquidity risk
When modelling sources of liquidity it is important to allow for factors limiting the extent and speed of liquidity transfers within an organisation and between distinct entities (‘fungibility’). Such factors might be legal, regulatory or operational in nature.
State key considerations when designing a set of appropriate (liquidity) scenarios and list some specific scenarios that should most likely be considered by a financial services company
©
Specific scenarios
Consideration should be made of:
1. both short-and long-term scenarios
2. appropriate interactions between risks – particularly between liquidity, market and interest-rate risks
Specific scenarios for consideration by a financial services company:
1. rising interest rates
2. ratings downgrade
3. large operational loss
4. large single insurance claim or a large set of associated claims
5. loss of control over a key distribution channel
6. impaired capital markets
7. sudden termination of a large reinsurance contract ©
Describe the two measures of liquidity risk introduced in Basel III ©
Measures of liquidity risk under Basel III
- Both measures categorise different types of assets according to their likely liquidity in times of market stress.
Liquidity Coverage Ratio (LCR) – designed to ensure that banks can survive a one-month stress scenario
LCR = stock of high quality liquid assets /total net cash outflow over the next 30 calendar days
Net Stable Funding Ratio (NSFR) – designed to consider funding over a one-year time horizon
NSFR = available amount of stable funding/ required amount of stable funding
Both ratios must be at least 100%. ©
State the possible components of insurance risk
Components of insurance risk
There are two main components of insurance risk:
demographic risk
non-life insurance risk.
Demographic risk can be broken down into:
1. level risk (or underwriting risk)
2. reserving risk, which includes:
a. volatility risk
b. catastrophe risk
c. trend risk (or cycle risk). ©
Define:
1. level (or underwriting) risk
2. volatility risk
3. catastrophe risk
4. trend (or cycle) risk ©
Level, volatility, catastrophe and trend risks
- level risk (or underwriting risk) – the risk that the particular underlying population’s claims incidence and intensity is not as expected over the short-term future, eg due to shortcomings in the underwriting process
- volatility risk – uncertainty with regard to the actual future immediate (short-term) claims experience (due to only having a finite pool of policies). As a consequence, going forward, the experience of sub-populations will exhibit statistical variations from that of the underlying population.
- catastrophe risk – an extreme form of volatility risk, eg the occurrence of a natural disaster resulting in a large number of claims
- trend risk (or cycle risk) – the risk of longer-term changes in claims incidence and intensity
©
Outline two methods of determining the current level of mortality ©
Determining the current level of mortality
1.
experience rating −examines the number of deaths in a homogeneous group of lives to determine the initial mortality rate or central mortality rate
2.
risk rating
−models the number of deaths (and hence the mortality rate) of each homogeneous group as a function of the shared characteristics of their members, eg socio-economic group. The model might, for example, take the form of a GLM.
−The shared characteristics of a population might be determined by proxy, eg postcode. However, all members of the group will not necessarily conform to the stereotypical risk factors upon which the model is then based.
These two methods may be combined by using credibility weighting factors
Describe how volatility risk might be modelled
Modelling volatility risk
Volatility risk can be modelled probabilistically or stochastically assuming some underlying statistical process.
As volatility risk varies by age models are generally fitted by Poisson maximum likelihood estimation (rather than by least squares optimisation):
1. using the model to be fitted, set the expected number of deaths at each age (a function of the parameters) equal to the mean of a Poisson distribution
2. calculate the probability of the observed number of deaths at each age based on the Poisson distribution (above)
3. maximise the likelihood function, ie the product of the probabilities (for all ages) to determine the fitted parameters
©
Describe how catastrophe risk might be modelled
Modelling catastrophe risk
The risk of a sudden, temporary increase in mortality (eg due to war or pandemic) is best modelled using scenario analysis – for example, considering a scenario under which there is a 20% increase in mortality at all ages.
More complex dependencies can be modelled by copulas. For example, it may be possible to consider multiple sources of mortality as separate risk factors each with their own probability distribution. These can then be combined using a copula into a multivariate distribution.
Note that catastrophe risk is one-way only. We can probably safely ignore the possibility of a sudden, temporary reduction in mortality.
©
Outline how other demographic risks (eg proportions married, lapse rates etc) might be variously allowed for when modelling
©
Assessing other demographic risks
Most other demographic factors (such as proportions married, numbers of children etc) are less likely to vary unexpectedly. These can be allowed for, when modelling liabilities, by using conservative assumptions for unknown independent variables.
Other demographic risks, such as lapse rates or pension scheme leavers, may be more significant. They are also often dependent on other risks – for example lapse rates may increase during an economic downturn as customers are unable to afford premiums. Such risks are best assessed using scenario analysis, due to the lack of historic data.
©
Outline the similarities and differences in approaches to assessing non-life and demographic risk
Assessing non-life insurance risk versus demographic risk
As with demographic risk, non-life insurance risk can be broken down into level and reserving risk (including volatility, catastrophe and trend or cycle risks).
In addition to the probability / frequency of claims there is the added complication of need to consider the severity (intensity) of claims.
Non-life risks can be further divided into two broad categories, depending on incidence rates: high frequency (eg motor) and low frequency (eg excess-of-loss) with greater variability in intensity.
©
Outline the challenges in assessing climate change risk
Challenges in assessing climate change risk
- The future may look very different to the past, so models that are calibrated using past data may give misleading results.
- There are multiple levels of uncertainty, eg about: −
a. the climate system itself and how it will respond to increasing atmospheric concentrations of greenhouse gas emissions
b. future levels of greenhouse gas emissions, which depend on the responses of governments, regulators, businesses and individuals, both to climate change targets and to technological and economic developments. - The uncertainty means a wide range of possible outcomes and economic impacts.
©
